Simple Guide to Drone CI with Docker Compose: A Real-World Example

Here’s a streamlined guide based on our working implementation:

kind: pipeline
type: docker
name: production-deploy

# Only run on prod branch pushes or manual triggers
trigger:
  branch:
    - prod
  event:
    - push
    - manual

steps:
  # Step 1: Debug & Environment Check
  - name: debug-info
    image: docker:dind
    volumes:
      - name: docker-socket
        path: /var/run/docker.sock
    commands:
      - pwd
      - ls -la
      - docker version
      - docker info
      - docker compose version

  # Step 2: Deploy Application
  - name: deploy
    image: docker:dind
    volumes:
      - name: docker-socket
        path: /var/run/docker.sock
    commands:
      # Show environment info
      - echo "Current working directory and contents:"
      - pwd
      - ls -la
      
      # Verify Docker setup
      - docker version
      - docker compose version
      
      # Stop existing deployment
      - docker compose down --timeout 30 --remove-orphans || true
      
      # Deploy new version
      - docker compose up -d --build --force-recreate
      
      # Show deployment status
      - docker compose ps -a
      - docker compose logs --no-log-prefix

  # Step 3: Tag Release
  - name: publish-release
    image: plugins/git
    settings:
      branch: prod
      remote: origin
    commands:
      - git tag -f "production-$(date +%Y%m%d-%H%M)"
      - git push origin --tags

# Required Docker socket access
volumes:
  - name: docker-socket
    host:
      path: /var/run/docker.sock

Key Features Explained

1. Simple Trigger System

• Runs on prod branch pushes
• Allows manual triggers for flexibility

1. Three Clear Steps

• debug-info: Verifies environment
• deploy: Handles the actual deployment
• publish-release: Tags successful deployments

1. Docker-in-Docker (dind)

• Uses docker:dind image for reliable Docker operations
• Mounts Docker socket for container management

1. Safe Deployment Process

• Gracefully stops existing containers
• Forces fresh builds to avoid caching issues
• Shows logs for debugging

1. Version Tracking

• Automatically tags successful deployments
• Uses timestamp-based tags for traceability

Usage

1. Setup

• Place this .drone.yml in your repository root
• Ensure Docker Compose file is present
• Configure Drone CI with your repository

1. Deployment

• Push to prod branch
• OR trigger manually from Drone UI

1. Monitoring

• Watch build progress in Drone UI
• Check logs for detailed status
• Verify container status after deployment

This simplified approach gives you a robust deployment pipeline without unnecessary complexity. It’s production-tested and handles the essential needs of a Docker Compose-based deployment.